
Mailing List Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[tlug] OT. Linguistic ?
- Date: Wed, 4 May 2005 07:48:42 +0000
- From: Mauro Sauco <sauco@example.com>
- Subject: [tlug] OT. Linguistic ?
- User-agent: Mutt/1.5.8i
For those who studied Linguistics:
I've been working on a project to develop a device that could
possibly help people with speaking disabilities.
A very simple approach, basically I made an input device that
interfaces with an interpreter, a database and a text2speech
engine (linux/python).
I got a working prototype and some candidates desperately waiting
to test it, however the input device currently has a limitation of
20,736 permutations (future words)
The question is: Are 20,736 words enough for making a "language"?
maybe I should ask, how many words do we use in the everyday
conversation? going to a supermarket, talking to friends,etc, nothing
too fancy. Can we get a good level of understanding with only 20,736
words ? (sorry if it's not clear, my English sucks)
input device 20,736 possible permutations (12(letters) ^ 4(word lenth))
|
interpreter
|
database
|
text2speech engine
|
ENGLISH (maximum 20,736 words)
JAPANESE (maximum 20,736 words)
is this limitation of 20,736 words a problem?
For those interested in helping, the idea is to put all this staff
into a PDA, possibly a Zaurus and make this as movable as possible.
What needs to be done: set a project, a web-site and a CVS
repository ; make an web engine for a collaborative updating of the
token/words database; improve/create the Japanese voices for
Festival/Festvox; Port the voices to CMUFlite; improve the input
device.
Provably this already exist, however I was surprised to find out
that no one in the JSRPD (Japanese Society for Rehabilitation of
Persons with Disabilities) have or knows such device.
I am trying to make something that is real, accessible and
inexpensive (a PDA, a memory and a couple of wiring).
Thanks for any help.
//mauro//
Home |
Main Index |
Thread Index